MBA Knowledge Library · Technology & AI

Technology & AI Practice and Review

Test your reasoning across the six-lesson Technology & AI foundation. Write an answer before revealing the worked explanation, and check assumptions, evidence, controls, and the practical consequences of each choice.

← Technology & AI hub · Business library

How to practice

Choose a topic and write your reasoning before opening the question. Use the lesson links to review concepts. These original fictional exercises do not save responses or calculate a score. Numerical answers need their units, assumptions, and limitations.

1. Choosing business software

Review the lesson →

Question 1: A tool has many features but cannot complete your required approval workflow. Is a low subscription price enough to choose it?

Worked answer: No. Test mandatory workflow requirements first. A low price cannot compensate for a requirement the process must meet. Include implementation, support, access, integration, and exit costs in the comparison.

Question 2: What should a realistic software trial demonstrate?

Worked answer: A representative end-to-end task, including required inputs, review, exceptions, output, and recovery. A polished feature demonstration is not evidence that your actual workflow works.

Question 3: Why evaluate data export and ownership before adoption?

Worked answer: You need to understand how records can be retrieved, reconciled, retained, and moved if the arrangement changes. An exit plan is easier to assess before the system becomes indispensable.

2. AI use cases and limits

Review the lesson →

Question 4: A draft is fluent and confident. What does that establish about accuracy?

Worked answer: Very little by itself. Trace material claims to appropriate evidence, check omissions and calculations, and assess the result against the task requirements.

Question 5: Does a second assistant agreeing with the first independently verify a claim?

Worked answer: No. Agreement is not primary evidence and can repeat the same unsupported assertion. Inspect actual sources or use another suitable verification method.

Question 6: Should a useful AI draft automatically become an external action?

Worked answer: No. Distinguish preparing a recommendation from authority to send, publish, edit records, or execute another action. Apply the intended workflow’s review and authorization.

3. Business process automation

Review the lesson →

Question 7: 300 monthly requests take four minutes each manually. The new process uses 240 one-minute reviews, 60 four-minute exceptions, and two monitoring hours. What is the net time release?

Worked answer: Manual time is 1,200 minutes, or 20 hours. New handling is 240 + 240 = 480 minutes, or eight hours, plus two monitoring hours. The modeled release is ten hours monthly.

Question 8: At $30 per hour, ten released hours are valued at $300. Tool cost is $90 monthly and setup is $1,050. What is simple value-based payback?

Worked answer: Net modeled value is $300 − $90 = $210 monthly; $1,050 ÷ $210 = five months. This does not prove cash savings or cash payback if the labor cost is unchanged.

Question 9: A task-creation request times out. Should the workflow simply retry until it receives success?

Worked answer: No. The destination may already have created the task. Use stable identifiers, appropriate duplicate protection, bounded retries, and reconciliation of uncertain outcomes.

4. AI output verification

Review the lesson →

Question 10: Twenty drafts each require five items. Reviewers find eight omissions across six drafts. Calculate item omission rate and draft completeness pass rate.

Worked answer: There are 100 required-item opportunities. Omission rate is 8 ÷ 100 = 8%. Fourteen drafts have no required-item omission, so draft completeness pass rate is 14 ÷ 20 = 70%. These are different measures.

Question 11: Two missing items are critical approvals. Can high overall readability cancel those failures?

Worked answer: No. Critical requirements should be explicit release conditions. Correct and recheck affected drafts rather than averaging the issue away.

Question 12: A draft calls a fall from 10% to 6% a 4% decrease. What should the reviewer clarify?

Worked answer: It is a four-percentage-point decrease, or a 40% relative decrease because (10 − 6) ÷ 10 = 40%. Preserve the intended meaning and denominator.

5. Spreadsheets, databases, and business systems

Review the lesson →

Question 13: An order total of $600 appears on each of three joined order-line rows. Is their $1,800 sum the order value?

Worked answer: No. The parent order total has been repeated at the line grain. Aggregate at the appropriate grain or use the correct line values.

Question 14: Does a valid foreign-key relationship prove the order belongs to the right real customer?

Worked answer: No. It can establish that a referenced record exists under the defined constraint. Correct real-world matching still needs appropriate evidence and checks.

Question 15: Does a transaction in one database automatically roll back an email already sent by another service?

Worked answer: No. Transaction guarantees apply within their defined boundary. External actions need their own confirmation, reconciliation, and recovery design.

6. Implementation and adoption

Review the lesson →

Question 16: A migration expects 960 records and receives 960 rows. Has it passed reconciliation?

Worked answer: Not on count alone. Check unique identifiers, expected values, relationships, exclusions, and source samples. One missing record and one duplicate can leave the same count.

Question 17: Twenty users attend training, eighteen sign in, and fifteen independently complete the task correctly. Calculate the three rates.

Worked answer: Attendance is 100%, sign-in is 90%, and demonstrated task completion is 75% of the invited cohort. These measure different things; investigate why task completion falls short.

Question 18: A pilot releases three minutes on 400 monthly tasks but needs six added support hours. What is net modeled capacity?

Worked answer: Gross release is 1,200 minutes, or 20 hours. Subtract six support hours to get 14 hours monthly before other omitted effort. Check comparable work, quality, and whether the result persists.

Review your reasoning

Identify whether each missed exercise concerns workflow fit, authority, evidence, denominators, record grain, or recovery. Revisit the related lesson and explain the answer in your own words before trying again. The exercises are general business education rather than a recommendation to change any live system.

Review the lessons

Examples, calculations, and recommendations are original educational illustrations, not measured product results or universal standards. Adapt decisions to the actual workflow, evidence, and approved information-handling requirements.